# Rethinking Monocular Depth Estimation with Adversarial Training

arXiv: https://arxiv.org/abs/1808.07528 · HTML: https://mlsanity.com/paper/1808.07528/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 22 Aug 2018.

## Built on by (most cited first, 5 of 5)

- [Colonoscopy 3D Video Dataset with Paired Depth from 2D-3D Registration](https://mlsanity.com/paper/2206.08903/) (2022)
- [Deep Learning-based Depth Estimation Methods from Monocular Image and Videos: A Comprehensive Survey](https://mlsanity.com/paper/2406.19675/) (2024)
- [Gated3D: Monocular 3D Object Detection From Temporal Illumination Cues](https://mlsanity.com/paper/2102.03602/) (2021)
- [Gated2Gated: Self-Supervised Depth Estimation from Gated Images](https://mlsanity.com/paper/2112.02416/) (2021)
- [RealNet: Combining Optimized Object Detection with Information Fusion Depth Estimation Co-Design Method on IoT](https://mlsanity.com/paper/2204.11216/) (2022)

## Abstract (arXiv)

Monocular depth estimation is an extensively studied computer vision problem with a vast variety of applications. Deep learning-based methods have demonstrated promise for both supervised and unsupervised depth estimation from monocular images. Most existing approaches treat depth estimation as a regression problem with a local pixel-wise loss function. In this work, we innovate beyond existing approaches by using adversarial training to learn a context-aware, non-local loss function. Such an approach penalizes the joint configuration of predicted depth values at the patch-level instead of the pixel-level, which allows networks to incorporate more global information. In this framework, the generator learns a mapping between RGB images and its corresponding depth map, while the discriminator learns to distinguish depth map and RGB pairs from ground truth. This conditional GAN depth estimation framework is stabilized using spectral normalization to prevent mode collapse when learning from diverse datasets. We test this approach using a diverse set of generators that include U-Net and joint CNN-CRF. We benchmark this approach on the NYUv2, Make3D and KITTI datasets, and observe that adversarial training reduces relative error by several fold, achieving state-of-the-art performance.